This, that, and the other form of Aβ have shown themselves harmful to neurons in various settings, but debate continues to rage over which are most menacing in Alzheimer's disease. A paper in this week's Journal of Neuroscience describes a way to determine relative toxicities of Aβ species in vivo. Developed by Luc Buée of INSERM and the University of Lille, France, and colleagues, the approach involves repeated hippocampal injections of synthetic Abta oligomers into freely moving mice—causing tau hyperphosphorylation, neuron loss, and memory deficits. While the authors say their method provides an in vivo readout for evaluating Aβ preparations and potential therapeutics, other scientists for now remain unconvinced of the pathophysiological relevance of the model.

Small, soluble Aβ oligomers—ranging in size from dimers to dodecamers—have been fingered as key drivers of neurotoxicity (Lambert et al., 1998; Lesné et al., 2006; Shankar et al., 2008; Ono et al., 2009). "We wanted to find a way to test all these Aβ oligomers in a controlled animal model," Buée told Alzforum (see also Benilova et al., 2012 review). Analyzing transgenic mice is tricky because "you don’t know which oligomers you will have," he said. Injecting Aβ into the mouse brain is another approach, but past research suggests that anesthesia alone sends phospho-tau levels soaring (ARF news story on Planel et al., 2007).

With these issues in mind, first author Jonathan Brouillette and colleagues set about injecting Aβ into awake, mobile mice. They prepared synthetic Aβ oligomers using a protocol from co-author Bart De Strooper’s lab at the University of Leuven, Belgium (Kuperstein et al., 2010), and injected them through surgically implanted tubes into the hippocampal dentate gyrus of 12-month-old B6 females once a day for six days. The injected material was primarily low-molecular-weight Aβ42 of less than 30 kDa—including monomers, dimers, trimers, and tetramers—as revealed by Aβ immunoblotting (6E10 and 3D6), and by transmission electron microscopy, atomic force microscopy, plus various spectroscopy methods.

Analyzed 24 hours later, the mouse brain showed Aβ oligomers crowding the injection site. It also showed extensive neuron loss, reduced levels of the NMDA receptor subunit NR2B, and elevated levels of cleaved caspase-3. Controls injected with vehicle solution and scrambled Aβ appeared normal. In addition, tau phosphorylation levels rose in the brains of Aβ-injected mice, though only at the S202/T205 site and not at other sites typically affected in AD. If the Aβ oligomers were pre-incubated with the sequestering agent transthyretin (TTR) prior to injection, the mice were spared neuron loss and memory deficits. Previous work suggested TTR protects against AD pathology in transgenic mice (see ARF news story on Buxbaum et al., 2008).

"With this model, we can test which forms of Aβ are most toxic," Buée said, noting it should also be useful for testing therapeutic approaches.

Other scientists wondered how well the data reported to date apply to AD. "The relevance of synthetic Aβ to the disease state is questionable," John Cirrito of Washington University School of Medicine, St. Louis, Missouri, wrote in an email to Alzforum (see full comment below). "It would be interesting to see what naturally derived Aβ oligomers would do in this chronic infusion system."

Moreover, the study used non-physiologic concentrations of Aβ; specifically, an amount that would reach micromolar concentrations if the injected oligomers distributed uniformly throughout one side of the forebrain, according to Karen Ashe and colleagues at the University of Minnesota, Minneapolis (see full comment below). "In comparison, Aβ dimers isolated from AD brains are cytotoxic at sub-nanomolar concentrations (Jin et al., 2011)." The authors used high Aβ concentrations to induce toxic oligomer conformations, and said that doing so accelerates processes that would otherwise take decades—too slow to permit laboratory studies. "Such high Aβ concentrations might actually be quite relevant for what happens in vivo," countered co-author Iryna Benilova of the University of Leuven. "It is suggested that Aβ can concentrate in intracellular compartments, thus creating the conditions for local formation of potentially toxic aggregates." (Hu et al., 2009)

The study did not address the possibility that the oligomers could have formed larger or smaller aggregates after injection. "I would be curious to know whether the Aβ oligomers that wind up in the hippocampus, and can be seen immunohistochemically, go on to become fibrillar (by electron microscopy)," said Lary Walker of Emory University in Atlanta, Georgia. The study did not test if other oligomers besides Aβ could have caused the neuron loss and memory symptoms.

Despite reservations expressed by some researchers, "we believe the authors should be lauded for their careful characterization of the Aβ species present in their preparations before injection and, impressively, the Aβ species actually present in the brain after injection," wrote Hsiao and colleagues.—Esther Landhuis

Comments

  1. This is an interesting study from Luc Buée’s and Bart De Strooper’s groups reporting the effects of synthetic Aβ1-42 oligomers (prepared as mentioned by Kuperstein et al., 2010, another study from Dr. De Strooper) on neuronal loss and cognitive function using a model of chronic injection in vivo. In this work, the authors report that the repeated daily injection over 6 days of apparent synthetic low-molecular weight Aβ oligomers (LMW oAβ) leads to neuronal toxicity as evidenced by FluoroJade staining, increased detection of active caspase-3 immunoreactivity, and decrease levels of GluN2B. In parallel, murine tau is hyperphosphorylated at S202/T205 while other sites traditionally affected in the human disease are not. Importantly, mice that received injections of LMW oAβ displayed apparent cognitive deficits in the passive avoidance memory task and in the Y-maze. The study is nicely complemented with applications of the same Aβ mixture onto primary mouse neurons to show that the same abnormal changes observed in vivo are also seen in vitro. Lastly, the authors use transthyretin in an attempt to alter the oligomerization/fibrillization process of Aβ and thereby limit oAβ-induced toxicity. As hypothesized by Brouillette and colleagues, transthyretin coincubation with synthetic LMW oAβ attenuated tau hyperphosphorylation and memory deficits in vivo and caspase-3 mediated toxicity in hippocampal neurons.

    While the data are clearly presented and the experiments well designed in general, I am puzzled by the overall conclusion. I do not fully understand how the effects observed are attributed to LMW oAβ. While the authors go to great lengths in carefully describing the biochemical properties of their preparations at the start of the experiment (day 0), I did saw no evidence that these alleged synthetic LMW oAβ species remain as such following 6 days of injections. In my opinion, this is problematic in such settings because as the authors wrote in their introduction, Aβ species "dynamically continue to oligomerize during that time." This point is well illustrated in Fig.1. How do the authors know that the effect observed is not due to larger Aβ assemblies such as dimer-based protofibrils (O’Nuallain 2010), annular protofibrils (Kayed et al., 2009 /pap/annotation.asp?powID=84759) or even fibrillar Aβ species formed within the last 5 days of their experiment? The same comment can be made for the in vitro paradigms following the application of synthetic oAβ for 3 days. It would be very interesting to see whether antibodies such as A11, OC, and α-APF (created in Charlie Glabe’s laboratory) would reveal the evolution of these initial LMW Aβ oligomers over time.

    Related to the nature of the Aβ molecules used in this study, I also struggled to fully understand the study’s interpretation of TTR as a sequestering agent of oAβ. In these experiments synthetic Aβ is incubated for 24h, leading to the formation of larger Aβ molecules referred as high molecular weight (HMW) oligomers based on their size (>35 kDa by SDS-PAGE) as well as fibrillar Aβ species (Fig.1B,C). Increasing amounts of TTR led to apparent reductions in fibrillar Aβ (Fig.6B, right panel) and increased oligomeric and monomeric Aβ levels (Fig.6B, left panel) as assessed by SDS-PAGE. It is possible that these oligomeric/monomeric Aβ molecules are bound to TTR, but there are no data demonstrating that is the case. Based on the figures presented, one could advance that TTR treatment favored the formation of Aβ oligomers at the expense of fibrillar Aβ. This could be possible as TTR (depending on its own conformation) was shown to disrupt fibril formation (Du & Murphy, 2010) as discussed by the authors. In that scenario, the neuroprotective effects observed in presence of TTR should be attributed to the ability of TTR to prevent fibril/protofibril-induced toxicity, not low-n Aβ oligomer-induced toxicity. Not knowing whether these peptides incubated for 24h (Aβ and TTR + Aβ) were ultracentrifuged at high speeds (i.e. 100,000 x g for 60 min) to remove fibrillar material from the preparation also makes it challenging to definitely conclude that the overall effects observed here are due to LMW oAβ oligomers.

    Despite the latter points, the efforts to characterize the Aβ peptides used in this study by Brouillette and coworkers should be applauded. In my personal view, especially when using amounts of Aβ peptides that are abundant enough to be readily detected (in the μM range), it is crucial to determine the nature of these assemblies shortly prior and shortly after the experiment to ensure that interpretations and conclusions about specific Aβ entities are as accurate as possible.

    References:

    . Characterization of the interaction of β-amyloid with transthyretin monomers and tetramers. Biochemistry. 2010 Sep 28;49(38):8276-89. PubMed.

    . Annular protofibrils are a structurally and functionally distinct type of amyloid oligomer. J Biol Chem. 2009 Feb 13;284(7):4230-7. PubMed.

    . Neurotoxicity of Alzheimer's disease Aβ peptides is induced by small changes in the Aβ42 to Aβ40 ratio. EMBO J. 2010 Oct 6;29(19):3408-20. PubMed.

    . Amyloid beta-protein dimers rapidly form stable synaptotoxic protofibrils. J Neurosci. 2010 Oct 27;30(43):14411-9. PubMed.

    View all comments by Sylvain Lesne
  2. Brouillette and colleagues have developed a system where they can inject synthetic Aβ oligomers into mouse brain, then detect toxicity and accumulation of the exogenous Aβ. There are several instances in the literature of synthetic Aβ being more toxic and less stable than naturally-derived Aβ oligomers. Consequently, the relevance of synthetic Aβ to the disease state is questionable. It would be interesting to see what naturally-derived Aβ oligomers would do in this chronic infusion system. A growing literature has found naturally-derived Aβ usually alters synaptic plasticity but rarely suggests that these Aβ oligomers are toxic. That would seem to contradict this report. The repeated infusion of Aβ oligomers into brain, as used here, is an important advance; however, I have strong doubts about the relevance of what is being infused.

    An open question within the AD field is, when are Aβ oligomers present in the brain? Are oligomers physiologic or only pathophysiologic? To answer those questions we will likely need animal models that endogenously express Aβ. Or even better might be a careful examination of Aβ species throughout aging in human brain and fluids.

    View all comments by John Cirrito
  3. Buée, De Strooper and colleagues have presented a novel mouse model in which to study Aβ toxicity in vivo. Six daily injections of synthetic Aβ oligomers into the dentate gyrus of normal adult mice led to transient tau hyperphosphorylation and pronounced local neurodegeneration. This observation contrasts with observations in multiple lines of APP transgenic mice, in which very little or no neuron loss is seen. Why this new model differs so drastically from transgenic models of Aβ toxicity is an important question, and bears on the relevance of the animal models to the human disease.

    First, the Buée study employed very high concentrations of Aβ, in an effort to accelerate pathogenesis. (If the injected Aβ distributed uniformly throughout one side of the forebrain, we estimate that it would reach micromolar concentrations; in comparison, Aβ dimers isolated from AD brains are cytotoxic at sub-nanomolar concentrations (Jin, M., et al., 2011) It is not at all certain that such a rapid response to a high concentration of a toxin is equivalent to the response to chronic exposure to low levels of the toxin. Second, work in our lab has shown that the in situ compartmentalization of specific Aβ oligomers profoundly influences their neurological effects (“in situ” refers to the location of Aβ species in the brains in which they are actually produced). For example, the rTg9191 mouse created in our laboratory generates ~1 µg Aβ dimers per forebrain (~450 nM if uniformly distributed), but these dimers are sequestered around plaques and their toxic effects are restricted to the immediate vicinity of plaques (Liu, P., et al., 2011). Therefore, exogenous administration of Aβ into the brain of a host animal might cause neurological effects that differ from those of even the same Aβ species in situ.

    Third, Buée has taken a cellular toxicology study usually done in culture and put it into a mouse brain. Is this like plaque-associated cytopathology spread over a wider area and without the plaque? This is a much more difficult system to work with than a cell culture system, and its only conceivable advantage is that one can look at effects on old neurons in situ as opposed to embryonic neurons in a dish.

    Despite our reservations about the relevance of this new model to AD, we believe that the authors should be lauded for their careful characterization of the Aβ species present in their preparations before injection and, impressively, the Aβ species actually present in the brain after injection. They have thus set a new standard for documenting the aggregation state of exogenously administered Aβ proteins.

    References:

    . Soluble amyloid beta-protein dimers isolated from Alzheimer cortex directly induce Tau hyperphosphorylation and neuritic degeneration. Proc Natl Acad Sci U S A. 2011 Apr 5;108(14):5819-24. PubMed.

    View all comments by Kathleen Zahs
  4. Some of the authors' observations are not surprising, from our view, since they performed an in-vitro incubation of Aβ oligomers with transthyretin (TTR), and we (and others) have shown that such an incubation will inhibit Aβ fibril formation, oligomer formation, and oligomer-induced cytotoxicity in tissue culture (see Buxbaum et al., 2008, and Li et al., 2011). In essence, the authors are doing an in-vitro inhibition of aggregation measured with an in-vivo neuronal readout. This is similar to Dominic Walsh’s oligomer inhibition of LTP in hippocampal slices.

    Technically, I might mention that the authors have no controls for TTR or Aβ. Do the oligomers form larger or smaller aggregates after injection? Would putting β2 macroglobulin or islet amyloid polypeptide oligomers into the dentate gyrus do the same thing? Would any TTR inhibit the damage or would another protein thought to “chaperone” Aβ have the same effect? The authors could have used clusterin or a TTR that does not have similar inhibitory activity in vitro.

    Although they correctly quote Du and Murphy, they identify TTR as an Aβ “sequestering” protein. There are no data indicating that TTR sequesters Aβ in vivo, since no one has shown that non-toxic Aβ TTR aggregates accumulate anywhere in either humans or transgenic mice. We showed in our PNAS paper that there is less Aβ40 and 42 extractable from the brains, not that the same amount is there and is non-toxic, which would have supported a “sequestration” hypothesis.

    References:

    . Transthyretin protects Alzheimer's mice from the behavioral and biochemical effects of Abeta toxicity. Proc Natl Acad Sci U S A. 2008 Feb 19;105(7):2681-6. PubMed.

    . Neuronal production of transthyretin in human and murine Alzheimer's disease: is it protective?. J Neurosci. 2011 Aug 31;31(35):12483-90. PubMed.

    View all comments by Joel Buxbaum
  5. This is a fascinating study from Luc Buée’s and Bart De Strooper’s groups reporting the effects of Aβ oligomers on neuronal loss and reduction levels of the NMDA receptor subunit NR2B, and elevated levels of cleaved caspase-3. This observation contrasts with other observations previously reported in various transgenic mouse models of AD, in which very little or no neuron loss is seen. The most important questions in this report are, What type or size of Aβ oligomers (ranging in size from dimers to dodecamers) causes neuronal loss, and does the degree of neuronal loss vary? Have the authors investigated the role of Aβ40 oligomers on neuronal loss in this particular mouse model? Finally, how relevant is this mouse model to the human disease?

  6. Many advantages can be attributed to this novel, flexible in-vivo approach:

    • The nature of toxic Aβ intermediates can be more accurately controlled by injecting Aβ preparations that are characterized before and after chronic injection, as we did in our paper (Fig. 1 and Fig. 2D).
    • Since the intrahippocampal injections are done in awake, freely moving mice, there are no confounding interference effects between any anesthetic agents and the Aβ solution on intracellular pathways.
    • To take into account aging—the most robust risk factor associated with AD—the effects of soluble Aβ1-42 oligomers were determined during the process of aging in 12-month-old mice. Chronic Aβ1-42 injections can also be done in younger and older mice to see their effects at different ages.
    • The collateral injection of soluble Aβ1-42 oligomers and vehicles permitted the control of any alteration within the same mouse.
    • Since Aβ accumulates in a time-dependent manner, the number of injections and the dose of Aβ can be adjusted to obtain more or less severe readouts of Aβ pathogenicity.
    • Because cell death occurs in the proximity of the Aβ injection site, neuronal loss can be induced in various and very localized brain regions.
    • The toxic effect of Aβ oligomers on molecular and cellular pathways can also be determined before and after neuronal loss within a reasonably short timeframe.
    • Since Aβ species were cleared gradually after injection, the long-term effects of Aβ oligomers after their removal can be analyzed both at the molecular and behavioral levels.
    • This new animal model can be used for preclinical validation of agents designed to prevent Aβ neurodegeneration, as shown in our paper using transthyretin (TTR).

    As discussed in the manuscript, TTR monomers have previously been shown to bind more extensively to Aβ monomers, impeding the further growth of Aβ aggregates (Du and Murphy, 2010). On the other hand, TTR tetramers interact more with Aβ aggregates than with Aβ monomers, and have been observed disrupting fibril formation (Du and Murphy, 2010). Thus, one could argue that the neuroprotective effect of TTR is mainly caused by the prevention of fibril/protofibril-induced toxicity. Although we cannot completely exclude the possibility that part of the neuroprotective effect of TTR is attributed to this mechanism, we think that the major mechanism for the TTR-mediated protection against Aβ toxicity is the sequestration of discrete toxic species and the arrest of Aβ monomer growth into multimers, since we observed that solutions containing an elevated concentration of small Aβ species were more toxic than Aβ1-42 preparation containing larger oligomers (Fig. 5).

    In summary, our novel animal model recapitulated many key neuropathological hallmarks of AD in a time-dependent manner, such as Aβ accumulation, marked neuronal loss, abnormal tau phosphorylation, and memory dysfunction. This in-vivo approach can prove useful in determining the toxicity of Aβ preparations as a function of their temporal profile.

    Since current methods of oligomer characterization are very limited and provide only semi-quantitative information, it is difficult to compare different oligomeric preparations in terms of concentration, conformation, and their potential relevance to the disease. In terms of oligomer identification and characterization, "in-vitro" and "ex-vivo" Aβ preparations have their own advantages and pitfalls (for more details, see our critical review, Benilova et al., 2012). In our study, we used recombinant human Aβ42 that was previously assessed under denaturing and non-denaturing conditions using Western blot, transmission electron microscopy (TEM), atomic force microscopy (AFM), Fourier transform infrared spectroscopy (FT-IR), electrospray-ionization mass spectrometry (ESI-MS), and nuclear magnetic resonance spectroscopy (NMR) (Kuperstein et al., 2010; Broersen et al., 2011). In future studies, it will be interesting to determine if Aβ preparations isolated directly from AD brains can induce similar effects using this model.

    References:

    . The toxic Aβ oligomer and Alzheimer's disease: an emperor in need of clothes. Nat Neurosci. 2012 Jan 29;15(3):349-57. PubMed.

    . A standardized and biocompatible preparation of aggregate-free amyloid beta peptide for biophysical and biological studies of Alzheimer's disease. Protein Eng Des Sel. 2011 Sep;24(9):743-50. PubMed.

    . Characterization of the interaction of β-amyloid with transthyretin monomers and tetramers. Biochemistry. 2010 Sep 28;49(38):8276-89. PubMed.

    . Neurotoxicity of Alzheimer's disease Aβ peptides is induced by small changes in the Aβ42 to Aβ40 ratio. EMBO J. 2010 Oct 6;29(19):3408-20. PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Anesthesia and AD: Phospho-Tau Surges in Sleeping Mice
  2. Fighting Fire With Fire—Transthyretin Therapy for Aβ?

Paper Citations

  1. . Diffusible, nonfibrillar ligands derived from Abeta1-42 are potent central nervous system neurotoxins. Proc Natl Acad Sci U S A. 1998 May 26;95(11):6448-53. PubMed.
  2. . A specific amyloid-beta protein assembly in the brain impairs memory. Nature. 2006 Mar 16;440(7082):352-7. PubMed. RETRACTED
  3. . Amyloid-beta protein dimers isolated directly from Alzheimer's brains impair synaptic plasticity and memory. Nat Med. 2008 Aug;14(8):837-42. PubMed.
  4. . Structure-neurotoxicity relationships of amyloid beta-protein oligomers. Proc Natl Acad Sci U S A. 2009 Sep 1;106(35):14745-50. PubMed.
  5. . The toxic Aβ oligomer and Alzheimer's disease: an emperor in need of clothes. Nat Neurosci. 2012 Jan 29;15(3):349-57. PubMed.
  6. . Anesthesia leads to tau hyperphosphorylation through inhibition of phosphatase activity by hypothermia. J Neurosci. 2007 Mar 21;27(12):3090-7. PubMed.
  7. . Neurotoxicity of Alzheimer's disease Aβ peptides is induced by small changes in the Aβ42 to Aβ40 ratio. EMBO J. 2010 Oct 6;29(19):3408-20. PubMed.
  8. . Transthyretin protects Alzheimer's mice from the behavioral and biochemical effects of Abeta toxicity. Proc Natl Acad Sci U S A. 2008 Feb 19;105(7):2681-6. PubMed.
  9. . Soluble amyloid beta-protein dimers isolated from Alzheimer cortex directly induce Tau hyperphosphorylation and neuritic degeneration. Proc Natl Acad Sci U S A. 2011 Apr 5;108(14):5819-24. PubMed.
  10. . Amyloid seeds formed by cellular uptake, concentration, and aggregation of the amyloid-beta peptide. Proc Natl Acad Sci U S A. 2009 Dec 1;106(48):20324-9. PubMed.

Further Reading

Papers

  1. . Neurotoxicity of Alzheimer's disease Aβ peptides is induced by small changes in the Aβ42 to Aβ40 ratio. EMBO J. 2010 Oct 6;29(19):3408-20. PubMed.
  2. . A specific amyloid-beta protein assembly in the brain impairs memory. Nature. 2006 Mar 16;440(7082):352-7. PubMed. RETRACTED

Primary Papers

  1. . Neurotoxicity and memory deficits induced by soluble low-molecular-weight amyloid-β1-42 oligomers are revealed in vivo by using a novel animal model. J Neurosci. 2012 Jun 6;32(23):7852-61. PubMed.